Traditionally, antigen-specific T-cells have been generated by selective expansion of peripheral blood T-cells natively specific for the target antigen. However, it is difficult and quite often impossible to select and expand large numbers of T-cells specific for most cancer antigens. Gene-therapy with integrating vectors affords a solution to this problem as transgenic expression of Chimeric Antigen Receptor (CAR) allows generation of large numbers of T-cells specific to any surface antigen by ex vivo viral vector transduction of a bulk population of peripheral blood T-cells.
Chimeric antigen receptors are proteins which graft the specificity of an antigen binder, such as a monoclonal antibody (mAb), to the effector function of a T-cell. Their usual form is that of a type I transmembrane domain protein with an antigen recognizing amino terminus, a spacer, a transmembrane domain all connected to a compound endodomain which transmits T-cell survival and activation signals (see FIG. 1A).
The most common forms of these molecules are fusions of single-chain variable fragments (scFv) derived from monoclonal antibodies which recognize a target antigen, fused via a spacer and a trans-membrane domain to a signalling endodomain. Such molecules result in activation of the T-cell in response to recognition by the scFv of its target. When T cells express such a CAR, they recognize and kill target cells that express the target antigen. Several CARs have been developed against tumour associated antigens, and adoptive transfer approaches using such CAR-expressing T cells are currently in clinical trial for the treatment of various cancers.
A number of toxicities have been reported from CAR studies, and additional theoretical toxicities exist. Such toxicities include immunological toxicity caused by sustained intense activation of the CAR T-cells resulting in a macrophage activation syndrome (MAS) and “On-target off-tumour” toxicity i.e. recognition of the target antigen on normal tissues. MAS is presumed to be caused by persistent antigen-driven activation and proliferation of T-cells which in turn release copious inflammatory cytokines leading to hyper-activation of macrophages and a feed-forward cycle of immune activation. A large spike in serum IL-6 is characteristic and the syndrome can result in a severe systemic illness requiring ICU admission.
On-target off-tumour toxicity has been reported with other CARs, for example a group of patients treated with a CAR against the renal cell carcinoma antigen CAIX developed unexpected and treatment limiting biliary toxicity. Two fatalities have been reported with CAR studies: one patient died of a respiratory distress syndrome which occurred immediately post-infusion of a large dose of 3rd generation anti-ERBB2 CAR T-cells; a further patient died in a different study after a possible cytokine storm following treatment of CLL with a second generation anti-CD19 CAR.
These toxicities are very difficult to predict even with detailed animal studies or non-human primate work. Crucially, unlike small molecules and biologics, CAR T-cells do not have a half-life and one cannot cease administration and wait for the agent to breakdown/become excreted. CAR T-cells are autonomous and can engraft and proliferate. Toxicity can therefore be progressive and fulminant.
A key aspect of these toxicities is immune over-activation i.e. over-activation of CAR T-cells.
Suicide genes are genetically expressed elements which can conditionally destroy cells which express them. Examples include Herpes-simplex virus thymidine kinase, which renders cells susceptible to Ganciclovir; inducible Caspase 9, which renders cells susceptible to a small molecular homodimerizer and CD20 and RQR8, which renders cells susceptible to Rituximab.
This technology adds a certain amount of safety to CAR T-cell therapy, however there are limitations. Firstly, it is a binary approach wherein all the CAR T-cells are destroyed upon addition of the suicide entity. In addition, medicinal therapeutics often have a therapeutic window. With a suicide gene the potency of the product cannot be tuned such that efficacy with tolerable toxicity can be achieved. Secondly, it is not clear whether a suicide gene would help with some of the immune-toxicities described above: for instance by the time a macrophage activation syndrome had been triggered, it may well no longer need the CAR T-cells to perpetuate and the suicide gene would no longer be helpful. The more acute cytokine release syndromes probably occur too quickly for the suicide gene to work. Finally, suicide genes are not “fail-safe”, i.e. the default status is for the CAR T-cells to be active.
There is thus a need for alternative methods for controlling CAR T-cells that are not associated with the disadvantages and problems mentioned above.